Evidence the run may read
- Expense lines and receipts
- Applicable expense policy
- Approved exception context
production blueprint · finance
Expense policy review agent blueprint.
Evaluate an expense packet against the applicable policy and prepare a cited decision for a human approver.
Search intent: AI agent for expense report review
Published by a2a cloudProduct-source methodology →
01 · the contract
Start from a bounded job, not a blank chat box.
trigger
An expense report enters policy review.
owner
finance owner
agent stops at
The designated finance approver decides reimbursement or exception treatment.
KPI · decrease
expense review touch time (minutes). Measure reviewer time spent per completed report.
outputs
Artifacts the run must produce
- Policy comparison
- Missing evidence list
- Approve, return, or escalate proposal
02 · topology
Small specialists. Named handoffs. One accountable decision.
Each stage produces an artifact another stage can inspect. The final node is a person, not an autonomous write to an external system.
- 01
intake
Validate and normalize the expense policy review inputs.
- 02
receipt matcher
Produce the policy comparison.
- 03
policy analyst
Produce the missing evidence list.
- 04
expense reviewer
Challenge the expense policy review result and prepare an approval packet.
- 05
finance owner
The designated finance approver decides reimbursement or exception treatment.
intake→ validated input packet →receipt-matcher
receipt-matcher→ Policy comparison →policy-analyst
policy-analyst→ Missing evidence list →expense-reviewer
expense-reviewer→ approval packet with evidence references →human-approver
03 · authority
Grant the run only what this case needs.
Source material is read-only. Drafts land in a case-specific output path. Tools may read or propose; the human gate owns the external write.
case inputs
workspace/finance/expense-policy-review/inputs/**Read only the evidence attached to this workflow instance.
case outputs
workspace/finance/expense-policy-review/outputs/**Write drafts and evidence artifacts without modifying source records.
approved tools
finance:expense-policy-review:read-or-proposeInvoke only tools explicitly granted for this run; external writes remain gated.
04 · implementation
A private, bounded starting manifest.
The blueprint starts private, caps its DAG, disables replanning, and exposes no public endpoint. Add only the tools and data adapters this workflow has approved.
a2a.yamlsafe starting point
name: finance-expense-policy-review
version: 0.1.0
entrypoint: agent:BlueprintAgent
expose:
public: false
composition:
planning: deterministic_dag
max_nodes: 6
max_parallel: 1
max_replans: 005 · acceptance test
Pass only with evidence.
- 01
Evidence traceability
Every material conclusion cites an input artifact or a scoped tool result from this run.
- 02
Every exception cites the exact policy rule and affected line; ambiguous items are escalated.
Every exception cites the exact policy rule and affected line; ambiguous items are escalated.
- 03
Approval boundary
The run stops at a proposal and records the human decision before any external side effect.
failure containment
Stop small. Preserve the evidence.
A receipt is unreadable or the applicable policy and jurisdiction cannot be established.
Containment: Return a partial result with unresolved items; do not broaden scope or perform an external write.
Operator: Attach the missing evidence, narrow the brief, or explicitly approve a new scoped run.
A required input or tool grant is unavailable.
Containment: Stop the affected branch and preserve completed artifacts in the case output workspace.
Operator: Grant only the missing resource or continue with that branch marked out of scope.
06 · proof
Sign the run facts. Keep money in the billing ledger.
Current platform receipts sign caller identity or classification, skill, bounded input evidence, verified grant IDs when present, outcome or result preview, and timing. Optional file, tool, artifact, handoff, evaluation, and review fields require separate instrumentation and are not populated by default. Price, fees, payouts, and later human approvals remain separate platform records.
The example uses only fields populated by the current platform sealing paths. It is illustrative, not a record of a real customer run.
ExecutionReceipt · selected fieldsEd25519 token
{
"receipt_id": "rcpt_01J...",
"schema_version": 1,
"agent_name": "finance-expense-policy-review",
"caller": "user:workflow-owner",
"task_id": "case_expense_policy_review",
"skill_name": "expense_policy_review",
"input_hash": "4d7c...9a2f",
"grant_ids": [
"grt_case_inputs",
"grt_tool_propose"
],
"status": "ok",
"result_preview": "Output prepared: Policy comparison. Human decision remains separate.",
"elapsed_ms": 4218
}related blueprints
Keep the operating model consistent.
Finance
Invoice exception review
Compare an invoice exception with purchase, receipt, and policy evidence and prepare a resolution recommendation.
Finance
Month-end reconciliation support
Compare ledger and subledger evidence, identify unmatched items, and prepare a reconciliation packet for sign-off.
Finance
Cash application review
Match a payment export to open receivables and prepare proposed applications with confidence and exceptions.
put the pattern to work
Start private. Scope the authority. Require the decision.
Deploy the workflow as a bounded internal agent, verify its outputs and the receipt fields actually emitted, then expand only the scopes your acceptance test proves it needs.